Icelab

Stephan Korsholm · Software consultant

AI makes writing code faster. But does it make your organization more productive?

AI coding assistants can dramatically accelerate code generation. Yet many software organizations struggle to translate that capability into faster, more reliable software delivery.

Let's discuss your AI development challenges
01

Better ways of working

The missing ingredient is often not better AI tools, but better ways of working with them.

I work with software organizations to adapt development workflows to exploit AI effectively.

02

From assessment to lasting change

I start by understanding how your organization works with AI today and identifying what is needed to realize its full potential.

In cooperation with both management and development teams, we then adapt your development practices, emphasizing well-defined, incremental changes, automated verification, and fast feedback.

03

Tools, infrastructure and results

I also implement the tools and infrastructure needed to support these practices. I work closely with your developers to ensure the new practices and infrastructure become a sustainable part of their everyday development process.

My goal is to turn faster code generation into faster, more predictable software delivery - without compromising quality.

I can also host and maintain the supporting infrastructure.

Experience behind the approach

Practical engineering. Real assignments.

My approach combines hands-on AI-assisted development with experience building the engineering practices and infrastructure that make reliable software delivery possible.

AI-assisted development

Modernizing a production Spring Boot application

Using Codex, automated tests, clear task specifications and code review, I completed around 15 changes alongside a framework upgrade from version 1.5 to 3.0 and authentication modernization. The assignment took about one month, compared with three months originally planned.

Read the assignment story on LinkedIn ↗

Development infrastructure

Building automated verification for an embedded team

I helped establish the automated development infrastructure for a new embedded software team within a large organization. The solution included automated builds and tests using both Linux-based simulations and physical hardware. An important part of the assignment was coordinating the needs of the development team with the organization's existing IT infrastructure and security requirements.

Read the assignment story on LinkedIn ↗